Grant: Discovery Grants Program - Individual
University of Toronto
$26,000.00
The Role of Neuromodulatory Systems in Meta-Learning
EF 2023-24
Résumé
Artificial intelligence can achieve remarkable feats including mastering the famously difficult games chess and Go. However, when it comes to the speed, flexibility, and abstraction of learning, humans still triumph. In this project, we want to understand what makes humans' learning abilities so unique by understanding the role that neuromodulatory systems play in meta-learning. Meta-learning refers to the ability to slowly learn abstract rules or constructs while also being able to swiftly adapt to changes in the learning context or environment. In this research program, we examine the roles of several neuromodulatory systems in the human brain, including dopamine, serotonin, acetylcholine, and noradrenaline, in orchestrating meta-learning. To achieve this, we propose the following innovations: (i) cognitive modelling: participants engage in a task where distinct aspects of uncertainty modulate learning and their behavior is explained individually with a computational model; (ii) advanced neuroimaging: brain activity is measured during the designed task using high-resolution functional magnetic resonance imaging (fMRI); and (iii) brain network modelling: the regional differences in brain activity are modelled as coupling strengths in a cognitive network. We hypothesise that distinct forms of uncertainty are represented in interacting brain networks, which map onto distinct neuromodulatory systems. One challenge is that these networks are governed by small structures in the brainstem, and non-invasive functional brain measurements with high signal-to-noise ratios and image resolutions of about 1 mm are required. In this work, we propose that cutting-edge technological advancements in fMRI can accomplish these requirements by rigorously optimizing the data acquisition method to capture the computational signatures of neuromodulatory systems. Research Aims: The aim of this proposal is to (i) optimise an advanced fMRI protocol for network identification of neuromodulatory systems, (ii) empirically validate competing network mechanisms in healthy controls with the advanced fMRI protocol and a pharmacological intervention study that aims to causally reduce the effects of each neuromodulatory system on learning. To achieve these aims, we will deploy previously developed candidate sequences for advanced fMRI at a cutting-edge 3 Tesla MRI system of the University Health Network in Toronto. Healthy human volunteers will perform a cognitive task, in which they have to elicit active ("go") or passive ("no-go") responses, while undergoing high-resolution fMRI. The mechanistic hypotheses of network-related contributions to meta-learning are tested by subject-specific computational modelling of task behavior and fMRI brain dynamics. Implications: Our project combines highly innovative ideas in both imaging and computational neuroscience. This unique combination may offer a breakthrough in understanding key neuromodulatory mechanisms supporting meta-learning.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
RGPIN-2022-04415
Name nom
Diaconescu, Andreea
Department département
Psychiatry
Organization id
31
Institution établissement
University of Toronto
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2023
Competition year année de concours
2022
Award amount
26000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1502
Committee name en
Biological Systems and Functions
Committee name fr
Systèmes et fonctions biologiques
Area of application code
1102
Area of application group en
Health, education and social services
Area of application group fr
Santé publique, éducation et services sociaux
Area of application en
Biomedical engineering
Area of application fr
Génie biomédical
Research subject code
2957
Research subject group en
Applied mathematics
Research subject group fr
Mathematical biology and physiology
Research subject en
Mathématiques appliquées
Research subject fr
Mathématiques de la biologie et de la physiologie
Application title
The Role of Neuromodulatory Systems in Meta-Learning
Keyword
active inference; brainstem; dopamine; functional magnetic resonance imaging; hierarchical Bayesian modelling; neuropharmacology; norepinephrine; reinforcement learning; serotonin; uncertainty
Application summary
Artificial intelligence can achieve remarkable feats including mastering the famously difficult games chess and Go. However, when it comes to the speed, flexibility, and abstraction of learning, humans still triumph. In this project, we want to understand what makes humans' learning abilities so unique by understanding the role that neuromodulatory systems play in meta-learning. Meta-learning refers to the ability to slowly learn abstract rules or constructs while also being able to swiftly adapt to changes in the learning context or environment. In this research program, we examine the roles of several neuromodulatory systems in the human brain, including dopamine, serotonin, acetylcholine, and noradrenaline, in orchestrating meta-learning. To achieve this, we propose the following innovations: (i) cognitive modelling: participants engage in a task where distinct aspects of uncertainty modulate learning and their behavior is explained individually with a computational model; (ii) advanced neuroimaging: brain activity is measured during the designed task using high-resolution functional magnetic resonance imaging (fMRI); and (iii) brain network modelling: the regional differences in brain activity are modelled as coupling strengths in a cognitive network. We hypothesise that distinct forms of uncertainty are represented in interacting brain networks, which map onto distinct neuromodulatory systems. One challenge is that these networks are governed by small structures in the brainstem, and non-invasive functional brain measurements with high signal-to-noise ratios and image resolutions of about 1 mm are required. In this work, we propose that cutting-edge technological advancements in fMRI can accomplish these requirements by rigorously optimizing the data acquisition method to capture the computational signatures of neuromodulatory systems. Research Aims: The aim of this proposal is to (i) optimise an advanced fMRI protocol for network identification of neuromodulatory systems, (ii) empirically validate competing network mechanisms in healthy controls with the advanced fMRI protocol and a pharmacological intervention study that aims to causally reduce the effects of each neuromodulatory system on learning. To achieve these aims, we will deploy previously developed candidate sequences for advanced fMRI at a cutting-edge 3 Tesla MRI system of the University Health Network in Toronto. Healthy human volunteers will perform a cognitive task, in which they have to elicit active ("go") or passive ("no-go") responses, while undergoing high-resolution fMRI. The mechanistic hypotheses of network-related contributions to meta-learning are tested by subject-specific computational modelling of task behavior and fMRI brain dynamics. Implications: Our project combines highly innovative ideas in both imaging and computational neuroscience. This unique combination may offer a breakthrough in understanding key neuromodulatory mechanisms supporting meta-learning.